FAUNA OF TUE AFUICAN LAKES. 603 



hazarded that the thalassoid Ciasteropod shells owe their nature 

 to the same cause. Be this as it may, there are additional com- 

 plicatious aflecting the salinity of Tanganyika, 



It has been shown that Tanganyika had probably no oiitlet 

 until a portion of the Nile basin became cut oft" and Kivu drained 

 south into the lake (p. 515). It has also been pointed out that 

 since Kivu water contains an excessive amount of magnesium 

 Sidts, that lake is probably the source from which the higli per- 

 centage in Tanganyika has been derived (p. 570). If these 

 probal)ilities be accepted, certain conclusions as to salinity follow, 

 louring the first period the salinity may well ha.ve been consider- 

 able, though there is no evidence as to its nature. The lake 

 subsequently freshened, but eventually its waters became rich in 

 salts of magnesium. Thus any effect which the saline nature of 

 the water may have exerted on the organisms of the lake may 

 have been due to either of these conditions, or to a combination 

 of both. As already suggested, there may even exist an irregu- 

 larity in the outflow of the lake, due to the forming and breaking 

 of dams in the bed of the Lukuga liiver (p. 515). If this be the 

 case, the salinity of Tanganyika has not only changed consider- 

 ably in the past, but may still be changing matei'ially from time 

 to time. 



The view that Tanganyika owes its remarkable organisms — not 

 merely the thalassoid forms — to a long-protracted period of 

 isolation, has been advocated bj^ several writers, and remains, 

 on the whole, the most likely suggestion put forward. The 

 possible effect of an increased salinity, which isolation would 

 involve, must of necessity be coupled with this, but it is not 

 regarded as the prime factor. This view has the positive 

 advantage that it does not run counter to geological conceptions, 

 but fits in with what is believed to be the past history of the lake. 



Testimony in favour of it is afforded by the very remarkable 

 nature of the Cichlid fishes which Tanganyika contains. This 

 group has long been known to show a peculiar facility for 

 colonising isolated and often saline watei'S, tliough the agency by 

 which this is effected is not understood. What then more likely 

 than that the Cichlids were among the earliest inhabitants of the 

 lake, where, without having to compete with other types of iish, 

 they multiplied unchecked and became differentiated into new 

 genera and species {cf. 26, p. 423). 



It is hardly necessary to point out that this isolation 

 hypothesis does not assume that Tanganyika was stocked from 

 any exceptional source. That is to say, the lake did not receive 

 its fauna from an ancient sea, but in the same manner as the 

 neighbouring fresh-waters, the original similarity of its fauna to 

 those of the other lakes being secondarily lost by marked 

 divergences of form consecpient upon prolonged isolation. The 

 marine aspect of certain Gasteropods would thus be regarded as 

 merely due to convergence. Viewed in this light, the case 

 of Tanganyika is closely analogous to that of oceanic islands, 



40* 



